Since neurofibrillary pathology is the first lesion observed in most AD cases (see Characteristics of AD), often preceding amyloid plaques, and deranged tau function can be mechanistically related to impaired microtubular function, a logical hypothesis was that altered tau was the primary etiologic agent in AD, with A^ as a secondary factor. The two competing dogmas attracted supporters and detractors, the "Tau-ists'' and the "^AP-tists." Although the sequence of histopathological changes supported the preeminence of tau, the genetics of the familial, early-onset forms of AD highlighted the importance of A^.
Tau transgenics The discovery of mutations in the tau gene in fronto-temporal dementia-17 (FTDP-17) familial tauopathy opened up the possibility that similar advantage could be taken of tau mutants to model pathology as had been so useful with ^APP. There was also the chance to settle the long-standing debate over the primary insult in AD. Although familial forms of various tauopathies were mapped to the tau gene, no tau mutations have been linked to AD, despite ample tau deposition in neuro-fibrillary tangles in AD. Tau is expressed as multiple splice variants in humans such that up to six forms are produced in developmental-, tissue-, and cell-specific patterns. Abnormal polymerization of tau is caused by mutations in coding or intronic regions of the tau gene, each somehow resulting in a characteristic tangle morphology causing a recognizable clinical phenotype resulting in different diseases such as Pick's disease, progressive supranuclear palsy, and multiple system tauopathy. Both the fibril morphology and dementia profile differ from those found in AD. Aft usually is not deposited in these diseases.
Unlike the case for mouse models of ft-amyloidosis, generating murine neurofibrillary tangles that mimic those in humans has been more challenging. Unlike humans, which express 3- and 4-repeat tau, mice produce only the four-repeat form. Overexpression of a number of proteins such as tau kinases increases the phosphorylation and accumulation of tau in axons, but the phenotype seldom resembles that seen in AD (Tesseur et al., 2000; Bian et al., 2002). Transgenic mice expressing human tau protein produce tangle-like structures in the CNS (Lewis et al., 2000; Gotz et al., 2004). When co-expressed with mutant human ftAPP transgenes (Lewis et al., 2001; Oddo et al., 2003a; Oddo et al., 2003b), tanglelike pathology is exacerbated in these mice.
The fibril morphologies in tau-transgenic mice resemble human tauopathy in the disease from which the mutant tau was derived, not the twisted organization of true NFTs that are common in AD. In most instances, tau in transgenic mice is expressed and deposited in areas where it is not seen in AD and yet is missing from areas in which it is found in AD (Lewis et al., 2001). Major pathology in these models occurs in the spinal cord and brain stem, with much less in cortical areas where the pathology is prominent in humans. Relative amounts of expression of the different splice variants seem to be part of the explanation. No Aft deposits have been detected in tau-only transgenic models. The current state of tauopathy models is summarized by Lee et al. (2005).
Combined Aft and tau pathology Critical evidence consistent with a role for Aft as a driving force in AD was obtained with mice doubly transgenic for ftAPP and for P301L tau (Lewis et al., 2001). Tg 2576 mice were crossed with JNPL3 mice carrying a four-repeat form of tau without the N-terminal repeats under control of the mouse prion promoter. Tau expression does not significantly alter regional Aft deposition patterns, while ftAPP expression results in tau deposition in AD-relevant areas where the tau is not deposited in the single tau transgenic animals. The morphology of the tau fibrils is not PHF-like, resembling, instead, the tauopathy from which the mutant tau had been derived. These observations and microinjection studies of Aft(1-42) fibrils into P301L tau-expressing mice (Gotz et al., 2001) indicate that Aft, and not tauopathy, is the primary insult in AD. Tissue culture models are being used to tease apart the complex biochemical network that connects the two pathologies (Gotz et al., 2004).
Triple transgenics (APP, PS1, P301L tau). Human ftAPP Swedish mutant K670N/M671L and human tau 4R P301L cloned separately behind neuron-specific Thy1.2 promoters were co-injected into the pronucleus of oocytes of homozygous PS1 M146V knock-in mice in which the human sequence I145V and human mutation M146V had been engineered into the mouse gene behind its natural promoter (Oddo et al., 2003a; Oddo et al., 2003b). The /APP and tau genes both incorporate at the same locus, greatly simplifying the breeding by ensuring that they are co-inherited. These animals develop synaptic dysfunction before plaques or tangles appear, and the deficits in LTP correlate with the accumulation of intraneuronal A/. The patterns and relative timing of A/ and tau deposition, which are distinct, closely resemble those in AD brain. While closer to AD-like tangle morphology, the tau deposits have not been definitively characterized as NFTs. Interestingly, A/ immunotherapy in these animals results in the clearance of amyloid plaques and of early, but not established, tau pathology (Oddo et al., 2004).
Was this article helpful?